| In recent years,Due to the continuous development of the world economy,industry not only brings great economic benefits to mankind,but also brings negative effects to the world that cannot be ignored.Carbon dioxide emissions are increasing year by year,posing a serious threat to the sustainable development of human beings.In the face of various threats brought by greenhouse gases,the development and utilization of renewable energy has become the main research work in the energy industry.Solar power is gradually being developed and applied in many fields.In solar power generation,factors such as illumination and temperature have a great influence on photovoltaic cells and affect the stability and continuity of the power generation system.Research and add energy storage system to solve the problem of photovoltaic power generation.In this paper,an independent photovoltaic dc microgrid with mixed energy storage is designed by using battery and ultracapacitor as energy storage components.The power distribution strategy of hybrid energy storage system is deeply studied,and the working mode of dc microgrid is analyzed.The main work of this paper is as follows.Firstly,the capacity configuration and characteristic analysis of photovoltaic power generation unit,hybrid energy storage unit and load unit in dc microgrid are carried out,and each module is modeled.The topology of DC/DC converter for each module connected to DC microgrid is analyzed.Secondly,the control strategy of the maximum tracking point of photovoltaic power generation is studied.The current and voltage of the photovoltaic power generation system are constantly subject to changes in the external temperature and radiation,resulting in constant changes in the maximum power point of photovoltaic power generation.To solve this problem,a new method based on self-coding neural network technology is proposed and applied to the maximum power point tracking of photovoltaic power generation.Based on deep learning network training of stacking encoder,this method USES the back propagation method with supervised learning to fine-tune the self-coding neural network.Finally,the model of photovoltaic power generation system is analyzed by MATLAB/Simulink.The results show that compared with the traditional incremental conductance method,this method can track the maximum power point of photovoltaic power supply more quickly and accurately,and improve the efficiency of photovoltaicpower generation.Then,the power distribution of hybrid energy storage system is studied,and the power distribution strategy of hybrid energy storage system is proposed.A hybrid energy storage control strategy and energy management algorithm based on low pass filter and fuzzy controller are proposed.Thirdly,the control strategy is implemented in the dc microgrid system.Then,the simulation model of dc microgrid system with composite energy storage was built in MATLAB/Simulink and verified by experiments.By comparing with the traditional algorithm,the effectiveness of the proposed control strategy was verified.Finally,a simulation experiment platform of optical-storage-charge-dc microgrid system is built.Secondly,the experiment platform of dc microgrid system was built by using photovoltaic simulator,lithium iron phosphate battery pack,programmable dc resistance load and other equipment,and the configuration software was used to design the PC upper computer system to verify the coordination strategy.In order to facilitate further study of the power supply system terminal products.This paper also USES TMS320F28335 as the control chip to make a small power bidirectional DC/DC converter prototype and tests its basic functions. |
PDF Full Text Request